Understanding the stages: the four planes of development

Maria Montessori did not divide childhood into stages because a curriculum needed tidy boxes. She divided it because she watched children, thousands of them, over decades, and found that the mind changes its entire mode of operation at fairly predictable points, not just its level of knowledge. She called these shifts the four planes of development, and the idea sits at the root of almost everything else in this method: why a Montessori classroom looks the way it does, why materials are introduced when they are, and why the same child who could not sit still at four might work in near-total absorption at five.

This page is a map of that territory. It draws on Montessori’s own late-life synthesis of the idea, most fully set out in The Absorbent Mind (1949, based on a 1946 lecture course) and The Secret of Childhood (1936), and on how later educators and researchers have used, tested and occasionally revised it since. Nothing below is a direct quotation; it is a plain-language account of the ideas, written for parents meeting them for the first time.

The shape of the pattern: building, then settling

Across all four planes, a pattern repeats: the first half of each roughly six-year period is a phase of intense, sometimes turbulent transformation, and the second half is a calmer period of consolidating what was built. The 0–6 and 12–18 planes are the turbulent ones; 6–12 and 18–24 are the calmer ones. Montessori thought this rhythm explained why early childhood and adolescence feel so disproportionately hard for both the child and the adults around them: they are, by design, the periods of greatest internal rebuilding.

0–3: the unconscious absorbent mind

In the first plane, Montessori argued, a young child takes in their surroundings (language, movement, the emotional tenor of the people around them, the rhythms of their culture) without effort, selection or conscious intention. She likened it to the way a sponge absorbs water: total, indiscriminate, and astonishingly fast. In her account, a baby raised among two languages does not “study” the second one; she takes it in from the people around her, because absorption at this age requires no deliberate attention.

This is also the window in which Montessori located what she called sensitive periods: windows of unusually intense, narrowly focused drive toward one specific kind of learning (language, or movement, for example), which fade once the relevant ability is established. In practice, this is why the 0–3 environment in a Montessori setting centres on what she termed practical life, meaning real, purposeful, scaled-down tasks such as pouring, spooning and dressing, rather than toys that merely simulate adult activity. The reasoning is that repeated, self-chosen, purposeful movement is how a very young child builds independence, concentration and a sense of order, not an incidental benefit alongside it.

3–6: the conscious absorbent mind

The same absorbency continues, Montessori held, but it becomes will-directed: the child starts choosing what to focus on, rather than absorbing everything indiscriminately. This is the age most people picture when they think “Montessori classroom”: child-sized furniture, open shelves of materials, children moving freely between activities.

A few structural features follow directly from the theory, not from tradition for its own sake:

Sensorial materials isolate a single quality (size, colour, texture, pitch) so the child can refine one sense of discrimination at a time, which Montessori treated as the groundwork for later abstract thought.
Auto-education: many materials are self-correcting, so a child can see their own error and adjust without an adult’s judgement intervening.
Freedom within clear limits: the child chooses their own work from a prepared range of options, inside a structure the adult maintains.
An uninterrupted work cycle, often around three hours, intended to let concentration build rather than be reset by the clock.
A three-year age span per classroom, so younger children can learn by watching older ones, and older children consolidate their own understanding by teaching it.

Around six, Montessori described a genuine shift in the child’s centre of gravity: away from the senses and toward reasoning, imagination and a developing moral sense, alongside a strong pull toward peer groups. She used the word “herd” deliberately, to capture how central belonging becomes at this age.

The signature idea of this plane is cosmic education: a single unifying narrative of the formation of the universe, the Earth, life, and human civilisation. It serves as scaffolding on which every subject (science, history, geography, maths, language) is hung, so the child builds one connected picture of the world rather than a set of disconnected subjects. This plane is also when Montessori introduced “going out” (structured excursions into the wider community to pursue a question that the classroom alone cannot answer), alongside a steady bridge from concrete materials toward abstraction, via timelines, charts and increasingly symbolic representations.

12–18: adolescence as a second reconstruction

Montessori treated adolescence as comparable in magnitude to the first plane: another period of intense physical and psychological rebuilding, this time layered with a need for social justice, economic self-sufficiency, and recognition as a capable person rather than a child.

Her own proposal for this age, Erdkinder (“land children” or “children of the land”), envisaged communities organised around shared, meaningful work on the land rather than a conventional timetable of lessons. It is, by a wide margin, the least developed part of her original programme. She wrote about it only briefly, in From Childhood to Adolescence, and most of what exists today under that name has been built out since by later practitioners and organisations such as NAMTA, extrapolating from a sketch rather than implementing a finished design. Worth knowing before assuming a 12–18 Montessori programme looks as settled as the 3–6 one: it generally does not.

What the research actually says

It is tempting, once a framework this coherent is laid out, to treat it as settled fact. It is not, and this project will not present it that way.

One major recent synthesis, a 2023 Campbell Systematic Review of 32 qualifying studies (Randolph et al.), found generally positive effect sizes for Montessori education, both academically and on measures like social skills and children’s own sense of their school experience, with larger effects in randomised than in non-randomised studies. Its searches ended in February 2020, so it could not include a national lottery-based trial published in 2025 and led by Angeline Lillard. That trial compared children who had applied for a place in a public Montessori preschool in the United States and were offered one by lottery with applicants who were not. At the end of kindergarten (ages five to six), it found effects on reading, executive function, short-term memory and theory of mind, but not on maths, vocabulary or several other outcomes; only 20.9% of families consented to take part, assessments were missing for 32% of the children offered a place and 42% of the others, and the results measure the effect of being offered a place rather than of attending. Evidence Status describes the trial in full.

At the same time, the evidence base is smaller and less standardised than for mainstream education research; the effect sizes in the 2023 review were smaller at middle and high school level and in public (as opposed to private) Montessori settings; and families largely choose Montessori rather than being assigned to it. Lottery studies take that choice out of the comparison, but they say most about families who applied, and low or uneven consent and missing data, as in the 2025 trial, can weaken them.

Researchers currently disagree about how strong this evidence is. A 2026 review reaches a sharply different conclusion about the Montessori evidence base from the 2023 review described above. It is a preregistered systematic review and meta-analysis (its plan was registered in advance) by León, Lipnevich and García-Garrido, published in the Review of Educational Research. It identified 198 studies, 114 of them with enough data for meta-analysis. It found that the pooled effects on learning outcomes were, at face value, positive and large by conventional benchmarks, both in studies that measured children before and after (pretest–posttest designs) and in studies that measured them only afterwards (posttest-only designs).

The authors judged those estimates not robust. The estimates showed extreme between-study heterogeneity (they varied enormously from one study to the next), substantial small-study effects (smaller studies tended to report larger effects) and clear signs of publication bias (a pattern suggesting that studies with positive results were more likely to be published). Inconsistent reporting of implementation fidelity (how faithfully schools actually followed the method), low transparency and limited experimental control further limit how far the findings generalise. The authors conclude that the large average effects reported in the research literature are largely explained by structural weaknesses in the evidence base rather than by robust educational impacts.

The review does not conclude that Montessori education has no effect. Two questions remain open: how far its studies overlap with those in the 2023 review, and whether either review’s estimate holds up under the other’s checks for bias. This site has not yet read the full text of the 2026 review, so everything said about it here comes from its published abstract. This is a live, unresolved disagreement between credentialled researchers, not a footnote: treat any confident claim about “what the research shows” on Montessori, including on this site, as provisional until it settles further.

In short: the 2023 review described above reported positive average effects; the 2025 trial found effects on some outcomes and not on others; whether the large pooled effects in the literature are robust is disputed; and Montessori education has not been shown to be better than every well-implemented alternative. Readers who want the full picture, including methodological caveats, will find it in Evidence Status.

Where this leads next

Understanding the four planes is the foundation for the rest of this section: what to actually look for when choosing a school (and which claims should make you cautious), and how to bring the same principles into Montessori at home without buying a houseful of specialist materials. For something free to try, Ten Quiet Minutes is a printable sheet for watching your own child at home.

Correction, 11 October 2026: an earlier version of this page said that effect sizes “tend to shrink or disappear” in middle and high school and in public settings, when the 2023 review found them smaller; described studies using admission lotteries as the strongest evidence in that review, and Angeline Lillard’s lottery-based research as the most frequently cited example of it, claims this site had not checked; said that the US Department of Education’s clearinghouse had been unable to draw firm conclusions, which this site had not checked either; described more engaged parents as an explanation that lottery studies cannot rule out, although a lottery compares applicants who were offered a place with applicants who were not, and the real limits of such studies are low or uneven consent and missing data; stated as fact Montessori’s own account of how a young child takes in a second language; and summed up the evidence as “well-supported for the early years in particular, promising overall”, a single verdict this site no longer gives. See the corrections log.

Correction, 10 October 2026: an earlier version of this page gave the wrong initials for two authors of León et al. (2026), omitted its DOI, and described it as re-examining the same studies as Randolph et al. (2023), as finding the earlier effects inflated and as criticising how transparently the earlier reviews reported their search strategies, which the published abstract does not support. It also called Randolph et al. (2023) the most substantial recent synthesis, although the 2026 review covers more studies. See the corrections log.

Sources consulted (paraphrased throughout; no text reproduced verbatim): Maria Montessori, The Absorbent Mind (English edition, 1949) and The Secret of Childhood (English edition, 1936), both subject to active translation copyright held by the Montessori-Pierson Publishing Company; From Childhood to Adolescence, on the second reconstruction and Erdkinder. Modern research: Randolph, J.J. et al., “Montessori Education’s Impact on Academic and Nonacademic Outcomes: A Systematic Review,” Campbell Systematic Reviews (2023); Lillard, A.S. et al., “A national randomized controlled trial of the impact of public Montessori preschool at the end of kindergarten,” Proceedings of the National Academy of Sciences (2025), https://doi.org/10.1073/pnas.2506130122; León, S.P., Lipnevich, A.A. and García-Garrido, M., “The Montessori Method Under Scrutiny: A Meta-Analytic and Metascientific Review for Greater Transparency in Education Research,” Review of Educational Research (2026), published online 7 July 2026, https://doi.org/10.3102/00346543261458396.